whatmodel.sh / GPT-4 (0314)

GPT-4 (0314)

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

GPT-4 (0314) is a closed-weights language model from OpenAI, released March 2023. Licence: proprietary.

Catalogue identity notes

Comparison model from the Epoch AI internal-evaluation CSVs (Epoch release date 2023-03-14). Dated API snapshot gpt-4-0314. OpenAI deprecations page: shut down 2026-03-26.

52.1±0.6comparison only: no current run option verifiedfrom 16 benchmark results across 10 sources

$ how it compares

The LLMs ranked either side of GPT-4 (0314), out of 828 with a score, then more from OpenAI.

#modelscorewhereRAMreleased
374Qwen2.5-Math 7B Instruct Alibaba52±2local~6GB2024-09
375LFM2.5 8B-A1B Liquid AI52±1local~6.5GB2026-05
376GPT-4 (0314) OpenAI52±1comparison only2023-03
377GLM-4 (0520) Z.ai52±1comparison only2024-05
378ERNIE 4.5 21B-A3B Baidu52±1comparison only2025-06
more from OpenAI
1GPT-6 Astra OpenAI87±1api $20.00/M2026-09
3GPT-6.1 Sol OpenAI85±1api $4.00/M2026-09
7GPT-6 Sol OpenAI83±1api $4.00/M2026-09

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▅▇█▇▅▄▃▁ #248 of 524 · 46% of its score · 4 tests

Beats Qwen3.5 4B, Qwen3-VL 32B Instruct, Solar Open 100B (102B-A12B) here, though they rank higher overall. Trails QwQ-32B-Preview, Ministral 3 3B Reasoning 2512 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEval88.4% ✎▁·▁▁▃▄▆▅█▃▁▁ beats 93/99easy · avg 79%▇▇▇▇▇▇▇ 17%sharp
HumanEval+79.3% ✎▁·▁▁▁▃▄▄▄█▃▁ beats 90/99easy · avg 74%▇▇▇▇▇▇ 15%sharp
CRUXEval (ZeroEval, 800 problems)74.5% ✎▁▂▆▃▆█▆▃▂·▁▁ beats 35/47medium · avg 61%▇▇▇ 8.1%sharp · at its level
Aider code editing (133 Exercism Python exercises, two attempts)66.2% ✎▁·▁▂▄▁▆▄█▆▂▂ beats 40/62medium · avg 54%▇▇ 5.8%at its level
chat & writing▁▁▁▃▅▄▆█▇▇▇▃ #294 of 463 · 23% of its score · 5 tests

Beats Mistral Small 4 (119B-A6.5B), Ministral 3 14B Reasoning 2512, Qwen2.5 Coder 32B Instruct here, though they rank higher overall. Trails Flan-T5 XXL, Granite 4.0 H 1B, Granite 4.0 1B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
AlpacaEval 2 length-controlled35.3% ✎▁▂▂▃▆▆█▆▄▄▁▁ beats 91/122medium · avg 39%▇▇▇▇▇ 14%at its level
Polyrating rating (bias-aware re-fit of Chatbot Arena votes, mid-2024)1213.0 ✎▂▁▂▁▃▇█▆▇▄▂▂ beats 85/95avg 1201▇ 3.1%blunt
LMArena (text)1287.8 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 118/379avg 1268▇ 3.1%blunt
MT-Bench (GPT-4 single-answer grading, LMSYS runs)8.9600 MT-Bench score (1-10) ✎▆▂▂▆▆▄██▂▄▄▆ beats 26/28avg 8.819 MT-Bench score (1-10)▇ 2.7%sharp · noisy
AlpacaEval 1.0 win rate (vs text-davinci-003, GPT-4 judge)94.8% ✎▂▂·▆·▇▃▆█▆▄▂ beats 40/45very easy · avg 96%▇ 0.2%sharp · maxed out · noisy
math▁▂▃▅▇▅▅█▄▁▁▁ #365 of 462 * · 6.0% of its score · 2 tests

Trails Gemma 7B (base), DeepSeek R1 Distill Qwen 1.5B, Tulu 2 DPO 13B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
MATH Level 5 (ZeroEval, 721 problems)26.1% ✎▁▁▁▆█▂▇▇▄··▂ beats 15/38hard · avg 34%▇▇ 4.6%sharp · at its level
OTIS mock AIME 2024–20250.6% ✎▃▃▄▃▃▃▇▇█▃▂▅ beats 3/193very hard · avg 8.5%▇ 1.4%sharp · noisy
other tests5 not in a task · 26% of its score
testresultvs the others in its rundifficultycountswhy
MMLU-Redux (ZeroEval, 2,778 questions)81.6% ✎▁·▁▁▄▄▄▅█▄▂▁ beats 34/53easy · avg 77%▇▇▇▇▇ 13%at its level
ZebraLogic (ZeroEval, 1,000 puzzles)27.1% ✎▁▁▃▆██▁▁·▁▁▁ beats 43/57hard · avg 18%▇▇ 5.6%sharp
GPQA Diamond35.7% ✎▁▁▁▁▂▄▆▄▅▇█▅ beats 29/208hard · avg 40%▇▇ 4.2%sharp
EQ-Bench 3 (emotional intelligence, judge Elo)984.7 ✎▁·▃▁▄▅▅▃█▄▁▁ beats 2/62avg 1044▇ 1.9%typical
EQ-Bench v1 (60 questions, paper run)53.4% ✎▂··▄·▂▆▄▂█·▂ beats 13/14medium · avg 57%▇ 1.4%at its level

$ where to run it

No verified API or download recorded yet · add one

$ how the score was worked out · EvalPlus / EvalPerf published… 32%, WildEval ZeroEval published r… 31%, AlpacaEval 1.0 and 2.0 leader… 14% and 7 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
EvalPlus / EvalPerf published evaluation runs2▇▇▇▇▇▇▇▇▇▇ 32%Apache-2.0
WildEval ZeroEval published results4▇▇▇▇▇▇▇▇▇ 31%Apache-2.0
AlpacaEval 1.0 and 2.0 leaderboards (tatsu-lab/alpaca_eval repository)2▇▇▇▇ 14%Apache-2.0
Aider code editing benchmark runs (old edit leaderboard)1▇▇ 6%Apache-2.0
Epoch AI — internal evaluations2▇▇ 6%CC BY 4.0
Polyrating: A Cost-Effective and Bias-Aware Rating System for LLM Evaluation (arXiv:2409.00696v3)1▇ 3%CC-BY-4.0
LMArena leaderboard dataset1▇ 3%CC BY 4.0
LMSYS Chatbot Arena leaderboard Space: MT-Bench column of the published leaderboard table1▇ 3%Apache-2.0
EQ-Bench 3 canonical leaderboard results1▇ 2%MIT
EQ-Bench: An Emotional Intelligence Benchmark for Large Language Models — authors' own runs1▇ 1%CC BY 4.0
prior · pull toward the middle▇ 0%
weighted average= 52.1

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json